Automatically Building VoIP Speech Parallel Corpora for Arabic Dialects

摘要

This article discusses the process of automatically building Arabic multi-dialect speech corpora using Voice over Internet Protocol (VoIP). The Asterisk framework was adopted to act as the main connection between the parties, for which two virtual machines were created: a sender and a receiver. The sender makes a VoIP call to the receiver using the Asterisk framework, while the receiver records the call automatically, a process that is repeated for all the audio files involved in the corpora. In this work, more than 67,000 automatic calls were made between the sender and receiver machines, generating VoIP Arabic corpora for four Arabic dialects. The resulting corpora can be considered the first Arabic VoIP parallel speech corpora and will be made freely available to researchers in Arabic NLP and speech recognition research.